Advancement in the Detection of Diabetic Retinopathy

The main objective of this book is to explore the early detection of Diabetic Retinopathy by accurately finding the presence of lesions like hard exudates and haemorrhages using novel image processing approach. This book investigates the severity level of Diabetic Retinopathy by detecting the presence of exudates. The input colour fundus images are preprocessed by a progressive Median filter and the exudates are segmented by applying hierarchical correlation clustering. Different features based on colour, brightness and dynamic shape are extracted. These feature vectors are applied as input to the Random Forest (RF) classifier. This classifier grade the severity of the disease as normal, moderate or severe based on the location of exudates with respect to the macula. This book also discusses an efficient localization of Optic Disc and macula by applying an image processing algorithm. The important anatomical structure, macula, is located with the help of Colour K-means segmentation algorithm. The detection of important symptom Haemorrhages in colour retinal images is also explained in this book.